When viruses were not in style: parallels in the histories of chicken sarcoma viruses and bacteriophages.

Studies in History and Philosophy of Biological and Biomedical Sciences 48 (2014) 189e199

Contents lists available at ScienceDirect

Studies in History and Philosophy of Biological and Biomedical Sciences journal homepage: www.elsevier.com/locate/shpsc

When viruses were not in style: Parallels in the histories of chicken sarcoma viruses and bacteriophages Neeraja Sankaran Ashoka University, Plot #2, Rajiv Gandhi Education City, Kundli, Haryana 131028 India

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 5 September 2014

The discovery that cancer may be caused by viruses occurred in the early twentieth century, a time when the very concept of viruses as we understand it today was in a considerable state of flux. Although certain features were agreed upon, viruses, more commonly referred to as ‘filterable viruses’ were not considered much different from other microbes such as bacteria except for their extremely small size, which rendered them ultramicroscopic and filterable. For a long time, in fact, viruses were defined rather by what they were not and what they could not do, rather than any known properties that set them apart from other microbes. Consequently when Peyton Rous suggested in 1912 that the causative agent of a transmissible sarcoma tumor of chickens was a virus, the medical research community was reluctant to accept his assessment on the grounds that cancer was not infectious and was caused by a physiological change within the cells. This difference in the bacteriological and physiological styles of thinking appears to have been prevalent in the wider research community, for when in 1917 Felix d’Herelle suggested that a transmissible lysis in bacteria, which he called bacteriophagy, was caused by a virus, his ideas were also opposed on similar grounds. It was not until the 1950s when when André Lwoff explained the phenomenon of lysogeny through his prophage hypothesis that the viral identities of the sarcoma-inducing agent and the bacteriophages were accepted. This paper examines the trajectories of the curiously parallel histories of the cancer viruses and highlights the similarities and differences between the ways in which prevailing ideas about the nature of viruses, heredity and infection drove researchers from disparate disciplines and geographic locations to develop their ideas and achieve some consensus about the nature of cancer viruses and bacteriophages. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Bacteriophages Tumor-inducing viruses Virus concept Thought styles Lysogeny parallel histories Reception

When citing this paper, please use the full journal title Studies in History and Philosophy of Biological and Biomedical Sciences

1. Introduction Bacteriophages and cancer viruses may not seem to have much in common at first glance, but the two groups of viruses have shared curiously parallel histories from the time of their discoveries in the 1910s until the 1960s when André Lwoff provided the scientific community with the first modern definition of a virus (Hughes, 1977; Lwoff, 1957; Van Helvoort, 1994b). Especially notable in the histories of cancer viruses and bacteriophages is the way in which the proposals by their discoverersdthat the causative agents of chicken tumors and an apparently transmissible bacterial lysisdmight be viruses (D’Herelle, 1917a; Rous, 1911, 1912), were

E-mail address: [email protected]. http://dx.doi.org/10.1016/j.shpsc.2014.07.012 1369-8486/Ó 2014 Elsevier Ltd. All rights reserved.

received by peers in their respective research communities. Although, the two discoveries were made by scientists working on opposite sides of the Atlantic on unrelated problems in different medical disciplines. The idea that either could be a virus was met with very skepticism in both cases (Becsei-Kilborn, 2010; Summers, 1999). In part, this resistance was due to the state of flux of the concept of virus during the early part of the twentieth century (Creager, 2002; Hughes, 1977; Van Helvoort, 1991; 1994b). Another important reason, this paper argues, was the prevalence of very different styles of research and thinking among different research groups interested in similar medical problems (Van Helvoort, 1993; 1994a). The achievement of consensus about the nature of viruses and linking of hitherto incommensurable research and thought styles led to an acceptance of the viral identity of these agents by researchers in different fields.

190

N. Sankaran / Studies in History and Philosophy of Biological and Biomedical Sciences 48 (2014) 189e199

Borrowing Plutarch’s device of parallel biography in his Lives, this paper juxtaposes the histories of the discovery and reception of the Rous Sarcoma virus (RSV) and bacteriophages to show that despite differences in the specifics of disagreement, the negative reception faced by Peyon Rous and Felix d’Herelle stemmed from common intellectual roots. Interestingly, although the research communities interested in cancer research and bacterial infections did not intersect or communicate with one another at the time, many of the arguments used against the viral identities of the cancer agents and phages were remarkably similar in their lines of reasoning. In both cases there were significant differences between the research and thought styles of the discoverers and those of their opposition, which led to nearly diametrically opposed interpretations of the same data. The discovery of the phenomenon of lysogeny in dysentery bacilli by a Belgian scientist posed a challenge, first to the idea of the bacteriophages as viruses (Bordet & Ciuca, 1921), and later more generally to the acceptance of the idea that different cellular effects could be caused by infectious agents that for generations of the host gave no other indications of their presence. Eventually it was the explanations provided in the 1950s by the French researcher Andre Lwoff about the nature of lysogeny that helped to resolve the issue and definitively identify both phages and the sarcoma agent as viruses. This paper traces the was in which lysogeny, or rather its resolution, became the lynchpin that brought about a convergence of thought styles on the nature of the causative agents of different phenomena such as tumor induction and bacterial lysis and established the criteria by which viruses are identified and demarcated from other disease agents such as bacteria today. 2. A tale of two discoveries and their reception 2.1. Peyton Rous and the chicken sarcoma agent In 1909, a worried farmer brought to the Rockefeller Institute for Medical Research in New York City, a Plymouth Rock hen which had “projecting sharply from the right breast, a large, irregularly globular mass,” (Rous, 1910, p. 697). At the time Rous, who had trained as a pathologist after obtaining his MD from Johns Hopkins Hospital in Baltimore, had just been appointed at the Rockefeller Institute by then director Simon Flexner, specifically to work on cancer research. Rous had taken up this position against the advice of his mentor W. H. Welch, who had warned him that “Whatever you do, don’t commit yourself to the cancer problem” (Andrewes, 1971, p. 64). Consequently, it is not surprising that although others at the Rockefeller reportedly showed little interest in the farmer’s problem, Rous seized upon the investigative opportunity as a means of vindicating his decision. The initial examination suggested that the tumor was sarcoma, a growth of connective tissue, bits of which, when transplanted either to other parts of the host or into the breast of an unaffected fowl of the same species, induced the formation of new tumors in these locations (Rous, 1910, p. 697). Rous’s conclusion from these initial studies was simply that “so far as tested, this avian tumour closely resembles the typical mammalian neoplasms that are transplantable” (Rous, 1910, p. 705). But, while earlier experimental efforts to transmit the tumors of such mammals as mice, rats and dogs using cell-free filtrates of tumor tissue had proven unsuccessful, the fowl sarcoma was transmissible by such means (Rous, 1911, p. 397). “The first tendency will be to regard the self-perpetuating agent active in this sarcoma of the fowl as a minute parasitic organism,” he therefore concluded, conceding however, that: It is conceivable that a chemical stimulant, elaborated by the neoplastic cells, might cause the tumour in another host and bring about in consequence a further production of the same

stimulant. For the moment we have not adopted either hypothesis, (Rous, 1911, p. 409). By the following year, however, Rous felt that he had gathered enough evidence to state quite definitively that although, “experiments with the chicken sarcoma have not yielded a method whereby a causative agent can be separated from the tumours of rats and mice [.] they clearly prove that the characteristics of malignant tumours in general are compatible with the presence of a living causative agent,” (Rous, 1912, p. 205). Among his reasons for believing in the parasitic nature of the etiological agent was the sustained ability of the cell-free filtrates to transmit cancer even after treatments such as drying, glycerination and successive cycles of freezing and thawing, all of which killed the tumor cells themselves (Rous, 1912, p. 204). His belief was further buttressed by the finding of two other chicken tumorsda bone tumor known as an osteochondrosarcoma (Rous, Murphy, & Tytler, 1912), and a second sarcoma distinct from the first (Rous & Lange, 1913)dwhich were also transmissible to new birds using cell-free filtrates of the tumor. “The findings with the chicken tumours largely demolish the theoretical basis on which objections to an extrinsic cause for cancer have been built up,” Rous concluded upon finding the osteochondrosarcoma, (1912, p. 1794), following up with an even stronger claim a couple of years later, that “It is perhaps not too much to say that their recognition [of the agents of these tumors] points to the existence of a new group of entities which cause in chickens neoplasms of diverse character” (Rous & Murphy, 1914, p. 68). In what is now an near-canonical account of the history of cancer and viruses, Rous’ claims and conclusions about the possible extrinsic, infective nature of the cancer agent were either rejected outright or met with considerable skepticism from the leading cancer-experts of the day (Andrewes, 1971; Becsei-Kilborn, 2003, 2010; Dulbecco, 1976; Van Epps, 2005). Rous’s discoveries came during a time of a growing consensus in the cancer research community that cancer was not an infectious disease and that its origins lay somewhere in the cell’s own inner mechanisms (Becsei-Kilborn, 2003, p. 2). His case was doubtless also hindered by the fact that he was unable to isolate any observable organism from any of these tumors, the morphologies of which also showed no evidence of infection, at least in any conventional sense. While he appears to have held fast to his notion of a viral cause for chicken sarcoma, Rous could not find similarly transmissible sarcoma or other tumors in any of the mammalian models. Consequently he turned his attention to other medical problems such as blood biochemistry, which perhaps seemed more urgent at the time (BecseiKilborn, p. 112). Although he eventually returned to the cancer problem, this later work focused on the Shopes papilloma virus, and Rous never worked with the sarcoma agent again after 1915. Studies on the chicken sarcoma, however, did not languish for long either at his home institution or abroad, a point I shall return to later in this essay. 2.2. Felix d’Herelle and the phenomenon of bacteriophagy At its outset, the story of Felix d’Herelle’s discovery of the bacteriophage bears little resemblance to Rous’ experience with the sarcoma agent. Whereas Rous, despite his maverick ideas, was very much part of the mainstream research establishmentdhaving trained at one of the most renowned medical schools in the United States e d’Herelle (1873e1949) was an outsider to the medical research community. Of French-Canadian or Quebecois origin, d’Herelle was never formally educated beyond high-school. He gained his knowledge of microbiology largely through selfinstruction, obtaining practical experience in a private laboratory he set up in his home in Montreal in 1897 (Summers, 1999, p. 5). He later gained recognition within scientific circles by working on a number of diverse problems for various scientific commissions in


Latin American and North Africa, notably on bacterial pathogens and the biological control (through these pathogens) of insect pests. In 1911 he became associated with the Pasteur Institutes, and did work for the branches in Algiers and Tunisia for some years before moving to the parent institution in Paris. There he began to work on bacterial dysentery, which, eventually led to his discovery of the phenomenon of bacteriophagy (Summers, 1999, pp. 11e47). The result of a bacterial infection that leads to debilitating gastrointestinal symptoms and even death, dysentery had become a matter of great urgency with the onset of the First World War. D’Herelle was assigned to investigate the disease in 1915 following an outbreak in the town of Maisons-Lafitte, outside Paris. The specific clinical presentation in this case had led the chief medical investigator Georges Bertillon to suspect that the Maisons-Lafitte outbreak was not caused by any of the hitherto known strains of the dysentery bacillus. D’Herelle quickly proved Bertillon correct through his investigations on the disease agent (D’Herelle, 1916). Subsequently he continued to study the bacillus, presumably to devise vaccine therapies against the disease (Summers, 1999, p. 49). He found that the cell-free extract of some of these cultures, while showing no signs of containing any live bacteria when viewed under an ordinary light microscope, possessed the ability to lyse or dissolve actively growing dysentery bacteria in other cultures. When some of the material from these dissolved bacteria (the bacterial lysate) was mixed with another batch of actively growing dysentery bacilli, the new bacteria also underwent lysis, thereby showing that the lysis was transmissible (D’Herelle, 1917a). Now, d’Herelle was not the first person to have reported transmissible lysis in bacteria. The first report of transmissible lysis has been published 2 years earlier by Frederick Twort, a medical researcher from London (Twort, 1915). Twort first observed the phenomenon, stumbled upon as it were, while trying to grow “filter-passing viruses” in artificial media. Although unsuccessful in growing viruses, during the course of these experiments Twort grew bacteria called micrococci from some smallpox vaccination material. Investigating these bacteria further, he encountered a hitherto unreported phenomenon, which he later labeled as “glassy transformation” (Duckworth, 1976, p. 794). He found that the micrococci underwent a lysis (thereby turning a turbid solution glassy) which was transmissible to new uninfected micrococci through the lysate. In the conclusion of his paper, Twort also mentioned the occurrence of a similar phenomenon in other bacteria, drawing special attention to the dysentery bacilli (Twort, 1915, p. 1243). In a manner somewhat reminiscent of Rous in his first sarcoma paper of 1911, Twort suggested a few possibilities as to the possible cause of the lysis, among them the “ultramicroscopic viruses” (Twort, 1915, p.). Unlike Rous, however, Twort did not render an entirely neutral judgment: “It is clear the the transparent material contains an enzyme,” he wrote, adding that, “the possibility of its being an ultra-microscopic virus has not been definitely disproved, because we do not know for certain the nature of such a virus” (Twort, 1915, p. 1242, emphasis in original). When d’Herelle reported his version of transmissible lysis two years later, however, he was unequivocal in his idea that the causative agent was a virus: I have shown that the disappearance of the dysentery bacillus coincides with the appearance of an invisible microbe endowed with antagonistic properties with respect to the pathogenic bacillus. This microbe, the true microbe of immunity, is an obligatory bacteriophage (D’Herelle, 1917b, p. 375. emphasis added).1

1

Translation by Summers (1999, p. 186).

191

D’Herelle’s first use of the term ‘bacteriophage’ in a descriptive sense rather than simply as a neologism for a new discovery is worth noting because it gives us insights into his thought processes about the cause and mechanism of the bacterial lysis. To him, the transmissible lysis was a manifestation of something that lived at the expense of these bacteria in much the same way as the bacteria themselves lived at the expense of their human hosts. As he would elaborate in his monograph, The suffix ‘phage’ is not used in its strict sense of ‘to eat’ but in that of ‘developing at the expense of;’ a sense that is frequently used elsewhere in scientific terminology. [.] This is precisely the interpretation to be given the term ‘phage’ in the word ‘bacteriophage’ (D’Herelle, 1922b, p. 21). As recounted in considerable detail by other historians, a protracted controversy over the nature of bacteriophage arose in the wake of d’Herelle’s announcement of his discovery. It lasted well into the 1930s and rippled outward from France to Belgium, then England and eventually, to the United States (Summers, 1999; Varley, 1986; Van Helvoort, 1992a; 1993). Whereas Rous had removed himself from the sarcoma virus picture within a few years of his discovery and only really resurfaced in 1966 to receive the Nobel Prize, d’Herelle remained enmeshed in the battles over the nature of bacteriophage throughout the early decades following his discovery. He steadfastly maintained his basic conception of the phage as bacterial parasite and devised different arguments to counter the many and varied challenges to his theory. Perhaps the biggest challenge to d’Herelle’s theory of the nature of the bacteriophage came from the discovery of the phenomenon of lysogeny by the Belgian scientist Jules Bordet a few years after d’Herelle’s announcement of his discovery (Bordet & Ciuca, 1921). Bordet had coined the term lysogeny to describe the spontaneous appearance of a transmissible lysis in certain cultures of bacteria growing in phage-free cultures. The lysate from these spontaneously lysed organisms were capable of causing lysis in later generations of bacteria. Bordet argued that it was impossible to imagine that the lysogenic bacteria could harbor viruses for generations without manifesting any signs of infection and then suddenly undergo lysis due to the action of those selfsame viruses. Rather, he suggested, the lytic principle in both bacteriophagy and lysogeny was some sort of enzyme or ferment produced by the bacteria themselves (Bordet, 1922). The Australian biologist Frank Macfarlane Burnet, who was for the most part a proponent of d’Herelle’s idea that bacteriophage was a virus, made a little headway in meeting the challenges to the theory posed by lysogeny (Burnet, 1929; Burnet & Lush, 1936), but a wider consensus on the matter of phage as virus was somewhat slower in forming (Sankaran, 2008; Van Helvoort, 1992a; Varley, 1981). There was some acknowledgment by certain groups of researchers regarding regarding the agent’s viral nature. The American Phage Group, for instance, treated the matter as a given from the outset of their discussions of the phenomenon, all but ignoring lysogeny for many years (Ellis & Delbrück, 1939; Stent, Watson, & Cairns, 1966). But the idea that the phage was a virus was by no means universally accepted until the 1950s. The acceptance of bacteriophage as a virus had to wait for a widespread consensus in the scientific community on its nature of virusduntil, as Lwoff put it, “viruses should be considered as viruses because viruses are viruses” (1957, p. 252). 2.3. Epochs of incredulity and belief More parallels may be discerned in the early reception to the idea that the agents of chicken sarcoma and bacteria were viruses. Based on a thorough examination of Rous’s published and unpublished papers, Eva Becsei-Kilborn has persuasively argued that Rous himself

192


painted a picture of marginalization and lack of acceptance of his ideas for the viral etiology of cancer (2010, pp. 113e114). Rather than the “downright disbelief” that Rous claimed (1967, p. 1921), she contends that it was the “specific characteristics of the development of scientific medicine” in the early decades of the twentieth century that made it difficult for Rous’ contemporaries “to incorporate the viral theory of cancer into mainstream cancer research” (BecseiKilborn, 2010, p. 115). A similar argument may be made in the case of the reception to the viral theory of bacteriophagy as well, the particular mainstream medical discipline in this latter case being bacteriology. At the time d’Herelle put forward his theory, bacteria were not yet recognized as true “genetic” organisms (Creager, 2007, p. 160; Sankaran, 2010, p. 589) and hence the idea that they could be infected by other microbes and suffer infection or disease themselves was somewhat difficult for most bacteriologists to accept. Also worth noting is the fact that the resistance faced by both Rous and d’Herelle to their ideas was by no means unanimous. While both men had their vocal detractors, their ideas also had their share of supporters. Although Rous distanced himself from work on chicken tumors after 1915, his ideas about their viral etiology did not remain fallow for long. A mere decade later, William Gye, a medical researcher in Great Britain not only confirmed Rous’ findings about chicken tumors but also revived the idea of the viral identity of the agent (Gye, 1925). The entry of the noted British virologist Christopher Andrewes into the debates soon thereafter (Gye & Andrewes, 1926) as well as Rous’ re-entry into cancer research through his work on the Shopes papilloma agent (Andrewes, 1971, pp. 648e651) ensured the revitalization of the matter of the viral etiology of cancer (Becsei-Kilborn, 2010, p. 142). Meanwhile, d’Herelle’s ideas about the viral identity of bacteriophage, while contested by many, never completely lost their currency. Burnet, who had become widely recognized as an expert on bacteriophages by the early 1930s2 had shown his sympathy for d’Herelle’s viewpoint almost from the outset of his involvement in the subject (Sankaran, 2008, pp. 88e90), perhaps most notably in his paper discussing phage growth (Burnet, 1929). Moreover, despite the continued opposition of Bordet, less than two decades after d’Herelle’s initial report, such noted researchers as the future Nobel laureate Max Delbrück were taking it for granted that the bacteriophage was a virus (Ellis & Delbrück, 1939). Notwithstanding such supporters, the more general attitude towards the viral identities for sarcoma and other cancer viruses as well as bacteriophages remained one of disbelief (Becsei-Kilborn, 2003; 2010; Van Helvoort, 1993; 1994a) and the next section of this paper attempts to contextualize these attitudes within the intellectual climate and state of knowledge of those times.

3. Why viruses were out of style in the early 20th century Lwoff’s description of virusesdadmittedly something of an insider joke that was, in his own words, “prosy, coarse and vulgar” (Lwoff, 1957, p. 252)dmay not be particularly informative about the nature of viruses, but the fact that he could get away with his statement at all underscores the fact that his audience shared common ground with regards to their understanding of what a virus was. A similar claim cannot made about the scientific community’s idea about viruses at the time of the discoveries of Rous

2 Evidenced by the fact that as early as 1926 as he recounted to his fiancée (Burnet, 1926), he was invited by the Medical Research Council (MRC) in England to contribute a chapter about the bacteriophages for a reference series on bacteriology (Burnet, 1930). He was also the author of a review (Burnet, 1934) that became widely regarded as a classic on the subject of bacteriophages.

and d’Herelle, the latter trained in the Pasteurian school of thought (Van Helvoort, 1992a, p. 255), which held that “Tout virus est un microbe [all viruses are microbes]” (Pasteur, 1890).3 Illana Löwy has pointed out that as historians of science, we should not forget that “many scientific terms do not possess a single, well-defined meaning” (Löwy, 1990, p. 88). The truth of this claim is evidenced by a comment to Rous in a letter from his friend, James Ewing, who was one of the strongest opponents of the idea of a living agent for sarcoma, “I presume it all settles down to the question, what is a virus” (Ewing, 1935, quoted by Van Helvoort, 2004, p. 201). In the early part of the 20th century viruses were “usually characterized by three negative properties” according to the leading American virologist, Thomas Rivers (1932, p. 423). As noted by historians Waterson and Wilkinson, the filterable viruses “were not retained by bacteriological filters; they could not be seen in the light microscope; and they could not be grown on artificial media” (1978, p. 78, emphasis added). Both Rous and d’Herelle’s early descriptions of their ideas about the phenomena they had discovered bears out the truth of the observation that these entities were better understood in terms of what they were not and could not do, rather than in terms of positively identifiable characteristics. Describing the causative agent of chicken sarcoma for instance, Rous reported, It will not pass through a dialyzing membrane, nor, in our experience, through a porcelain filter. These various features seem sufficient to identify it as a living organism in distinction from a ferment. The organism has never been directly observed in fresh or stained preparations; and the morphology of the individual tumor cells does not suggest its presence. Attempts to cultivate it in vitro have not as yet proven successful (Rous, 1912, p. 204; emphasis added). D’Herelle also employed similarly negative terminology, referring to the agent of bacteriophagy as “the invisible microbe,” reporting that the “antagonisitic microbe can never be cultivated in media in the absence of the dysentery bacillus” (d’Herelle, p. 158). Neither Rous nor d’Herelle used the actual term ‘virus’ in their early papers, preferring a host of such alternatives as, “living causative agent” (Rous, 1912, p. 205), “invisible microbe” (D’Herelle, 1917b) and “ultramicroscopic parasite of bacteria” (D’Herelle, 1922a, p. 292). But they did not voice any objections to the term when it was applied to their discoveries, and even used the term explicitly in later writings (D’Herelle, 1931; Kidd, Beard, & Rous, 1935). Indeed, on more than one occasion in later years, Rous indicated in letters to various colleagues that his omission was intentional (Rous, 1953, p. 98, 1958). His original impulse was to call the cause of the sarcoma tumor a ‘virus’ but was prevented from doing so by “an older and wiser man” (Rous, 1958), whom he also described as the “crusty, redoubtable, lovable old Secretary of the Board of Scientific Directors, DiPrudden, [who] put his granite foot down against it, suggesting ’agent’ instead” (Rous, 1953, p. 98). As Rous went on to confess, DiPrudden evidently [ . ] did me a good turn, since the virus proved in some ways so peculiar that not until the time of my Harvey lecture in 1934, when the traits of viruses generally were better realized could it safely be called as such (Rous, 1958). 3.1. Contrasting thought styles in bacteriophage research The flux in virus concept notwithstanding, it is clear from the responses of the opponents to both Rous and d’Herelle that it was the idea of an infective live agent rather than its specific identity as a

3

Directly translated as “Any virus is a microbe.”


virus to which they were objecting. In his analysis of some of the controversies over whether or not a bacteriophage was a virus, Ton van Helvoort has made a compelling case for interpreting them “as conflicts between competing research styles” (1994a, p. 98). He argued that the controversies were only resolved when the research styles were somehow linked to one another (Van Helvoort, 1992a, 1992b, 1993, 1994a). By ‘research style’da phrase he used interchangeably with ‘thought style’ e van Helvoort was referring to an concept introduced as an analytic tool in the social studies of science by the Polish physician and medical researcher Ludwig Fleck. According to Fleck, the work of scientists could be characterized and understood in terms of their thought styles (the closest translation in English, of what he called a Denkstil in German), which constitutes “the entirety of intellectual preparedness or readiness for one particular way of seeing and acting and no other” (Fleck,1979 [1935], p. 64). By Fleck’s reckoning, the thought style of a researcher was situated within what he called “a thought collective”dthat is to say a scientific community (Mößner, 2011, p. 417)dand was the driving force that both determined “the direction of research and connected it with a specific tradition” (Fleck, 1979, p. 64). In the case of the dispute over the nature of bacteriophage between d’Herelle and the Belgian microbiologist Jules Bordet in the 1920s, van Helvoort has argued that it was their different research styles which, “exerted a constraining force on what was to be studied and how this had to be done” (Van Helvoort, 1992a, p. 244). d’Herelle, who as indicated earlier, had stumbled upon bacteriophagy in the course of his research on dysentery, interpreted his findings in light of the Pasteurian tradition of regarding all infectious disease as being caused by living microbes (Van Helvoort, 1992a, p. 255). Consequently, when he observed bacterial lysis, d’Herelle would have compared the nature of damage to other known cases of cellular lysis and by analogy, would have attributed it to a similar cause, i.e. an infection of the bacterial cells. This style of thinking, which van Helvoort characterized as “bacteriological” (1992a) or “exogenous” (1994a) style, stemmed from a different context than that of Bordet. As an expert in the immunology against infectious diseases, Bordet viewed bacteriophagy in the context of bacterial variation (van Helvoort, pp. 253e254). Because Bordet conceived of variation as an expression of alterations in bacterial metabolism (physiology), van Helvoort dubbed his style as “physiological” (1992a, p. 253) or “endogenous” (1994a). A similar difference of thought styles also underpinned the differences of opinion regarding the modes of bacteriophage formation within the bacterial cell which arose between John Northrop and Max Delbrück in the 1940s (Van Helvoort, 1992b). Northrop, an enzyme chemist, saw the phages as endogenous enzymes produced by the bacteria and hence, their formation as a product of bacterial synthesis. Challenging this idea was Delbrück, a physicist turned molecular biologist whose primary interest was to use the bacteriophage as a tool to understand the working of genes at the molecular level. He therefore considered the phages as exogenous viruses and their production as a form of viral replication inside the bacterial cell (Van Helvoort, 1992b).4 3.2. The endogenous and exogenous thought styles in cancer research The opposition faced by both d’Herelle and Delbrück to their interpretations of bacteriophagy bears a striking resemblance to

4 Both Bordet and Northrop had won Nobel Prizes shortly before their involvement in the debates under discussion, and ironically, in the bacteriophage debates, both d’Herelle and Delbrück challenged them in the very areas for which each received his prize!

193

the arguments employed by various cancer researchers who disagreed with Rous’s interpretation that the chicken sarcoma was caused by an exogenous agent. A close examination of the work of Rous and d’Herelle working on chicken sarcomas and transmissible lysis of bacteria respectivelydreveals that despite their widely differing medical education and research training, the two shared a style of thinking and research rooted in their study of infectious pathology, broadly speaking. This similarity allowed both these men to consider exogenous agents or viruses as the culprits of the phenomena under their respective investigations. In his analysis of the reception of the oncogene paradigm, van Helvoort observed that the same dichotomy in thought and research styles that colored the bacteriophage debates was also the basis for the negative reception to the idea of what later came to called the oncogenic viruses (1999). The resistance to the viral theories of tumor formation for instance, was rooted in the same types of thought styles and traditions as the opponents of the phage-as-virus theories. Even the language and terminology used by researchers in the late nineteenth and early 20th centuries to describe cancer causation foreshadows the two thought-styles outlined by van Helvoort. For instance, at the annual Bradshaw lecture of the Royal College of Surgeons in 1903, which incidentally, marked the first time in nearly a decade that the topic of cancer was addressed in a public lecture at the college, the speaker grouped “the alleged causes of cancer formation into the entogenous (or intrinsic) and the ectogenous (or extrinsic). By entogenous causes are understood certain spontaneous and anomalous changes within the organism; by ectogenous causes those derived from outside the body” (Morris, 1903, p. 1505, emphasis added). The same dichotomy of opinions was also recognized at the second International Conference on Cancer held in Paris in October 1910: Throughout the whole scientific proceedings it was evident that there were two distinct schools of thought, representing respectively those who cautiously, and perhaps more vaguely than in the past, believe in a parasitic etiology, and those who regard such an etiology as quite incompatible with the natural history of cancer (The international cancer conference, 1910, p. 1267). As an interesting aside, some forty years later Andrewes identified the same line of division between the attitudes of different researchers toward the viral theory of cancer: “The virus theory has had its ups and downs, mostly downs, for pathologists in general have not regarded it favourably. On the other hand, many bacteriologists, and particularly virus workers, are impressed by arguments in its favour” (1950, p. 81). According to the veteran historian of medicine Edwin Ackerknecht (1958, p. 117), scientific studies of cancer had taken a “decisive turn” in the early part of the nineteenth century with the introduction of the microscope as a tool in pathological anatomy and the interpretation of findings of such studies in light of thendeveloping cell theories of life. One of the most active and famous exponents of this period was Rudolf Virchow, who considered all pathology in terms of cellular change and applied this angle of vision to his studies of tumors (Virchow, 1847). Virchow identified and classified different types of tumors into categories according to the tissues in which they originated and the types of cells they affected. For example, tumors of the connective tissue were classified as sarcomas. Tracing specific tumors in this way, to their tissues of origin, allowed investigators to conceive of tumors as originating within the cell (Becsei-Kilborn, 2003, pp. 24e 25). As one supporter of a parasitic theory of cancer causation commented, “The view that cancer arises from a cell which has become essentially altered and then has suffered from some peculiar irritation does not [.] shut out the view that this

194


alteration and irritation may be due to a parasite” (Plimmer, 1903, p. 1512). Paradoxical as it may seem then, Virchow’s theories, so steeped in the language of cancer’s cellular origins, also opened the door for “the search for a specific micro-organism of cancer” (Morris, 1903, p. 1505). The alignment of the early 20th century cancer researchers with either the exogenous or endogenous school of thought on cancer etiology bears out the idea that a thought style is colored, if not wholly determined, by the context in which research is conducted. As van Helvoort explained, these alignments were a “result of different analogies and different models which were felt to be different” (Van Helvoort, 1993, p. 14). In his Bradford lecture for instance, Morris, a surgeon with experience in operating on various tumors and hence familiar with the behavior of tissues, was emphatic in his opinion that cancer originated within the body. He thus favored the view that tumors were caused by the proliferation of embryonic cells in adult tissues and regarded “the “tumor germ” theory as “the mostdindeed, in my opinion, as the onlydprobably correct theory of cancer origin which has ever been put forward, as the only hypothesis which is consistent with all the known facts of cancer”” (Morris, 1903, p. 1510). The clinical experiences of surgeons as well as other physicians who saw and worked with actual cancer patients also predisposed them to not favor infectious explanations, as it was clear to them that cancer was not contagious through person-to-person contact or other means through which infectious diseases were normally spread (Becsei-Kilborn, 2003, p. 37). In contrast, researchers who appeared to favor extrinsic, microbial explanations for the origins of tumors (e.g. Park, 1898; 1903; Plimmer, 1903) were laboratory pathologists whose work required familiarity with the nature of infectious diseases. Conversant, for example, with the changes that occurred in tissues infected with different microbes (Plimmer, 1903), or with the epidemiology of infectious diseases (Park, 1903), these researchers drew analogies from their experiences to their observations of tissue samples from tumours. Accordingly they came up with exogenous explanations for these tumors. Despite the many theories that abounded within both the intrinsic and extrinsic camps, no concrete evidence seemed forthcoming about the causes for cancer at the beginning of the 20th century. Etiology remained a subject of much speculation and many hypotheses. The title of a 1905 address by E. F. Bashford, then Superintendent of Research of the newly founded Imperial Cancer Research Fund, asking “Are the problems of cancer insoluble?” gives us an idea of the status of research at the time (Bashford, 1905). The lack of evidence proved more damning for the exogenous thought style. For one, the failure to isolate any microbes with definitive links to tumors stood out in stark relief against the many microbes that were isolated and shown to cause specific diseases during the same period. Furthermore, the proponents of this school were unable to present a unified opinion on the nature of cancercausing agent. As one opponent to the parasitic theory observed: The many pathologists who claim to have discovered the cancer parasite, all are in disaccord as to its nature; thus, it has been ascribed to bacteria of many different kinds, to various protozoan parasites, to diverse kinds of fungi, and various other lowly parasites and pseudo-parasites (Williams, 1908, p. 229). With something to offer vis-a-vis cancer treatment through surgery and palliative care, then, those favoring endogenous theories of cancer causation held an advantage in being able to voice a consensus on the issue of etiology, which they publicly declared at the second International Conference on Cancer in 1910: In regard to the causation of cancer, a mass of evidence was presented supporting the view that cancerous tissue is really a

biological alteration of the tissue proper to the individual attacked by the disease, and that its peculiar properties may be explained without assuming the intervention of extraneous agencies, such as a hypothetical cancer virus (1911, p. 1008). Small wonder then, that when Rous put forward his viral hypothesis for chicken sarcoma a scant few months after this conference, the cancer-research community was reluctant to revive the debate. As they would have seen it, Rous was merely adding to the motley crew of extrinsic agents enumerated by Williams which had not been isolated, with no more evidence than before (1908, p. 229). Despite the charges of confusion and lack of cohesion that Williams leveled against the parasitic theories, however, the different endogenous theories were hardly more unified or cohesive. Indeed, Bashford had even grouped the hypotheses concerning cancer etiology into three main categories rather than two, separating the extant endogenous theories into two further groups of variants depending on ideas about tissue origin, while all exogenous theories were lumped together as one (1905, pp. 1507e 1508). Endogenous counter-theories to the bacteriophage were similarly varied although d’Herelle’s putative filterable virus was the only exogenous theory that was in play. Well over a decade after d’Herelle put forward his theory, Burnet was still able to identify no fewer than half a dozen hypotheses suggesting endogenous mechanisms for bacteriophagy in the introduction of a paper describing bacteriophage growth (Burnet, 1929, pp. 9e10). The only matter on which these theories appeared to concur was that whatever a bacteriophage was, it was not a bacterial parasite (Sankaran, 2006, p. 36). Looking back at the reception to Rous and d’Herelle’s conceptions of their respective discoveries, one can see that, like the properties by which viruses were characterized, the attitudes towards the idea that sarcoma agents and bacteriophages were filterable viruses were negative only “in relation to the paradigm of bacteriology, which interpreted infectious agents as autonomous living microbes” (Van Helvoort, 1994b, p. 186, emphasis in original). During the 1930s and 1940s, van Helvoort has argued, the concept of the filterable virus “was deconstructed” and the broader concept of virus maintained only because “because scientists stressed the similarities between viruses at the level of their pathological effect” (Van Helvoort, 1994b. p. 189). For example, Rivers, who had highlighted the negative properties of viruses in the first place, declared a preference for “a positive characterization of the virus, one emphasizing the intimate relationship that exists between them and their host cells” (Rivers, 1932, p. 423). This is not to say that the older criteria for defining viruses were abandoned altogether but rather, that the concept was expanded on the basis of other properties, and the meaning of ‘virus’ underwent variances through a convergence of exogenous and endogenous thought styles employed by various virus researchers. 4. Convergence and resolution: lysogeny as lynchpin By the 1930s even mainstream virus researchers such as Rivers had begun to consider the sarcoma agent as a virus, citing it as an example of the virusehost relationship, e.g. “the marked stimulation and destruction of cells induced by [viral] activity which [.] gives rise to tumors, such as Rous’ sarcoma” (Rivers, 1932, p. 423). There was a growing consensus, reflected in the following statement by a medical researcher that, “Experimental cancer research has shown that the concept of “the cause of cancer” is now obsolete” (Stowell, 1945, p. 286; emphasis in original). Furthermore, he added, cancer could have either intracellular or extracellular causes, enumerating examples such as “chemical carcinogenic agents, heredity, hormones, milk factors, viruses, physical trauma, and


precancerous conditions” as various contributory causes (Stowell, 1945, p. 286). Even this wider acceptance that viruses could be a cause of cancer, though never the cause, was insufficient to dispel doubts altogether. Rather than reflecting an unequivocal acceptance of the idea of the viral etiology of cancers, the terminology indicates the co-existence of multiple meanings of the word ‘virus’ at the same time, a state of affairs that lasted for some time to come. As late as 1958, for example, virologists still agreed that, “Presently the concept of viruses varies with the discipline of the investigators and the kind of viruses they studied” (Furth & Metcalf, 1958, p. 88). The tumor viruses in particular were set apart, at least in verbal references, from what were called the “classical” or infectious viruses, although as Furth and Metcalf conceded, the “evidence obtained on the morphology, biochemistry and antigenic makeup of tumor viruses has not indicated any essential differences” (1958, p. 90). Undoubtedly, a large part of the shift in attitude toward tumor viruses can be attributed to the “development of the new material culture” in the guise of such powerful technologies as electron microscopes, ultracentrifuges and crystallography which gave rise to a “new form of biological experimentation” (Creager & Gaudillière, 2001, p. 205). Working in concert with the influence of these developing technologies, I argue that a major conceptual driver for consolidating the consensus on the nature of the agents of chicken sarcoma and bacteriophagy was the resolution of the issue of lysogeny. While the influences of these different instruments and experimental techniques merit a deep historical analysis in their own right they lie beyond the immediate scope of this paper for the most part, and will only be touched upon as they pertained to clarifying the nature of phages or cancer viruses. 4.1. Refining resolution on the bacteriophages Lysogeny, as mentioned earlier, had been discovered in the early 1920s by Jules Bordet and for many years had posed the biggest conceptual hurdle to the idea that the bacteriophage, or agent of transmissible bacterial lysis, was a virus. Although it was not till 1950 that Lwoff “decisively resolved the debate” (Burian & Gayon, 1999, p. 325), others, notably, Eugéne and Elizabeth Wollman in France (Burian & Gayon, 1999, pp. 326e327; Galperin, 1987; Lwoff, 1953, pp. 276e280; ) and Burnet in Australia (Sankaran, 2008; Van Helvoort, 1992a, 1994a, 1994b) had begun to offer explanations as to how the lysogeny phenomenon might be reconciled with the notion of bacteriophage as an exogenous virus. Burnetdwhom Eugéne Wollman credited with being the scientist “to whom we owe the most complete study’ of bacteriophages and lysogeny” (Wollman, 1935, p. 1312)dwas probably the first person to provide a plausible explanation for the phenomenon. He described lysogeny as the “presence of bacteriophage or its anlage in every cell of the culture, i.e. it is part of the hereditary constitution of the strain” (Burnet & McKie, 1929, p. 282). Meanwhile Wollman had put forward a theory that “some genes could perhaps have a certain stability and be transmitted from cell to cell by the external medium” and suggested that the bacteriophages were perhaps such genes (Lwoff, 1953, p. 278; Wollman, 1928). Some years later he and his wife also recognized the existence of an alternation of infectious and non-infectious phases in the life cycle of bacteriophage (Wollman & Wollman, 1937). In a later paper on phage-induced mutation in bacteria Burnet explicated his ideas about the nature of the bacteriophage (virus) and lysogeny (gene) more explicitly: It is not possible to say whether this [mutation] results from an altered genetic constitution of the bacterium or is directly induced by the associated phage at each generation. According

195

to Wollmann’s hypothesis the distinction between the two alternatives would disappcar, the phage being regarded as a gene re-introduced into the genetic make-up of the organism (Burnet & Lush, 1936, p. 37). Burnet and the Wollmans’ explanations of lysogeny in terms of the gene was perhaps premature in the sense that it was a theory whose implications could not be linked simply and directly to the knowledge of the times (Stent, 2002). Regardless, by bringing the gene into play with their theories, they began to dissolve the hitherto sharply delineated boundaries between the exogenous and endogenous thought styles in virus research. Burnet and Wollmans’ ideas about lysogeny were in direct contrast to scientists such as Bordet, whose resistance to d’Herelle’s theory of bacteriophage on the basis of lysogeny was rooted in a belief in the rigid distinction between two modes of transmission; namely, the impossibility that a bacterium could carry a virus, i.e. an infection, for generations with no visible phenotypedplaque formationdand then spontaneously show signs of this infection for no obvious reason (Bordet, 1922). While the gulf between bacterial diseases and human disease might seem vast, it is worth remembering that d’Herelle described the process of bacteriophagy in terms that were explicitly analogous to human infection by bacteria (1917a, 1917b, 1924, p. 256). The Wollmans’ and Burnet’s insights into lysogeny underwent an eclipse, or in the words of the immunologist Melvin Cohn, “slept for twenty years to be awakened by André Lwoff whose Prince Charming kiss brought forth as progeny the fields of virology [etc]” (1978, p. 9). Among the reasons contributing to lysogeny’s long sleep were Burnet’s cessation of phage research after 1937 (Sankaran, 2010, pp. 593e594) and the tragic arrest, deportation and execution of the Wollmans by the Nazis in 1943 (Lwoff, 1953, p. 279). Scientific factors accounting for its neglect included the absence of a physical peg upon which to hang the theory as well as adequate ways to test its plausibility. Like the virus, concept, perhaps even more so, the concept of “gene” was in considerable flux, defined only in functional terms as the carrier of heredity. Nothing was known about the physical or chemical basis of either heredity or genes. To borrow the terminology of the sociologist of science Joan Fujimura, lysogeny for many years was not a “technologically doable” scientific problem (1987, p. 257). Perhaps nowhere is the state of flux in knowledge about such concepts as the gene, and virus made more evident than in the classic paper on bacterial transformation by Oswald Avery and his colleagues in which they first put forward the idea that DNA was the chemical carrier of heredity (Avery, MacLeod, & McCarty, 1944). In their discussion of the various interpretations of the phenomenon of transformation the authors named as alternatives virtually all the concepts and entities this paper has talked about thus far: The [transformation] inducing substance has been likened to a gene. [.] Another interpretation of the phenomenon has been suggested by [Wendell] Stanley who has drawn the analogy between the activity of the transforming agent and that of a virus. On the other hand, [James B.]Murphy has compared the causative agents of fowl tumors with the transforming principle of Pneumococcus (1944, p. 155, emphasis added). By the time lysogeny was reawakened in the 1950s much had changed in biology. By then Avery’s proposition that DNA was the material basis of genes was an accepted fact, as was the fact that viruses contained genes, which allowed Lwoff to replace “a traditional definition of viruses by a molecular one” (Morange, 2005, p. 592). Armed with a better understanding of the physico-chemical nature of viruses, Lwoff’s contemporaries were in a better position to visualize the process of lysogeny as he described it, and thus,

196


became more receptive to the notion of the dual existence of bacteriophage as both (or alternating between) a free virus and gene, or as Lwoff called it, the prophage. The explication of lysogeny brought along with it the modern concept of virus (still in place today), which in turn, enabled researchers to consider infective and lysogenic viruses as different forms of the same basic entity, thus dissolving the lines between the exogenous and endogenous styles once and for all. An important reason for this sea change was that the modern definition of a virus as an obligate intracellular parasites privileges the method of replication or multiplication as the main property that distinguishes viruses from bacteria: [.] contrary to bacteria, which multiply by binary fission, viruses go through an ‘eclipse’ when they reproduce themselves. In this process the virus particle loses its infectious character and is present only in the form of its nucleic acid, which codes for the hereditary information (Van Helvoort, 1994b, p. 190). In the case of the lysogenic viruses, the viral genome is integrated into the host cell’s DNA and hence becomes part of the cell, creating in effect a very long eclipse period in contrast to the nonlysogenic viruses, whose nucleic acid remains free and utilizes the host’s cellular machinery for replication as well as protein synthesis. 4.2. Cancer viruses and lysogeny: convergence by analogy Lwoff’s work on lysogeny and the prophages did indeed prove to be the “Prince Charming kiss” (Cohn, 1978, p. 9) that awakened the sleeping beauty that was virology. Generally speaking it opened up avenues of research by offering a new way of thinking about viruses and their relationships to their hosts, both at cellular and organismal levels. In the story of the phages and tumor viruses specifically, lysogeny proved to be the focusing lens through which the hitherto separate trajectories of research on the bacteriophages and cancer viruses converged, primarily serving as an exemplar of the way in which viruses could function as “genetic particles, interacting with or even integrated within the cell’s DNA” (Creager & Gaudillière, 2001, p. 228). Of course, as history shows us, there had been “a constant exchange of models and hypotheses between genetics, virology and oncology during the first half of the century” (2002, p. 425), as evidenced by the work of the aforementioned (Avery’s group 1944, p. 155) as evidence. But, it may be argued, lysogeny added something more to the picture, namely: “the possibility of an alternation between inactive and active phases that characterized both certain viral diseases and the evolution of cancers” (Peyrieras & Morange, 2002, p. 425). This fact evidently did not escape Lwoff’s notice either, for his review contains an extensive discussion of the analogy between the two situations (1953, pp. 323e 325). The implications of his ideas and the elegance of the model he provided were compelling enough to attract many researchers to explore the possible analogies of different plant, animal and insect viruses to the lysogenic bacteriophages. Furthermore, at a 1959 symposium dedicated to “The possible role of viruses in cancer,”5 the bearing of Lwoff’s molecular and genetic explanations for lysogeny on the cancer problem was the dominant topic of conversation.

5 Rous, who was one of the organizers of the symposiumdwhich brought together leading international researchers from different disciplines such as pathology, virology, microbial genetics and molecular biologyddescribed its main purpose as being “less to report new facts than to put forward our ideas and conclusions, and to deal with those of others, altering or accepting them straight off and doing our best for them and with them” (1960, p. 672).

Early in the symposium Andrewes, there as a self-described “general purpose virologist . ignorant about many relevant facts about genetics,” observed that: Dr. Lwoff’s word “prophage” has been a very useful one, and the corresponding one, “provirus,” directs our thought along very interesting channels. One of the most fruitful things that I think we shall have to discuss today is whether we are right in considering that viruses may go into a provirus state analogous to prophage. No one, so far, has produced any conclusive evidence that this is so. If it were, I think it would make sense (Andrewes, 1960, p. 689). These remarks were raised in direct response to the physicistturned-bacteriophage-researcher Salvador Luria’s talk which had dealt with “the place of virology in cancer research” (1960, p. 677). Identifying the key problem in cancer research more generally as the clarification of “mechanisms that control cellular growth and the cellular alterations that make the tumor cell deficient in internal controls or unresponsive to external ones,” (Luria, 1960, p. 677), Luria also pointed to the gamut of possibilities that bore investigation: “At one extreme, the virus may master-mind the whole process; at the other extreme, it may trigger only a single step” (1960, p. 678). “Instructive” as Andrewes found Luria’s presentation, his overall response is revealing about the conceptual gulfs that still existed between attendees who hailed from research traditions embedded in differentdexogenous vs. endogenousdthought styles. As an animal virologist, and hence favoring the extrinsic school of thought, Andrewes took particular exception to the molecular biology camp’s acceptance of the idea of viral infections of the cell as “a kind of infective heredity” (Luria, 1960, p. 679), a description clearly rooted in endogenous conceptions of cancer causation. “Generally accepted? This idea may or may not be a good one, but if he really thinks it is generally accepted it just convinces me that these geneticists never come out of their ivory towers. [.] When he says that ’virus infection is a cellular mutation,” I am personally at a total loss. I don’t think it even occurred to me during the 1957 epidemic of Asian influenza that we were suffering in our hundreds of thousands from cellular mutations! (Andrewes, 1960, p. 690). A further objection that Andrewes raised against the molecular biologistsdnot only Luria but also Lwoffdwas the implication that by integrating with the genome of the host cells, a virus was somehow losing its identity. As he recounted, After a long discussion in London in 1957, Dr. Lwoff came to the conclusion, if I remember right, that a virus was a virus because it was a virus. I absolutely agree with this. But I would go a little further: I think it is still a virus when it is what he calls a provirus. [.] I maintain that a virus integrated with the host’s genetic material and no longer separate is still a virus, just as a fly without wings is still a fly (Andrewes, 1960, p. 691). The above comments reflect just one instance of differences of opinion on a whole host of issues surrounding the main topic among the thirty-odd researchers who attended the symposium. At the same time, however, a perusal of the talks and discussions reveals that there was a clear sense of a growing consensus on certain central issues, chief among them that the genetic concept of virus infection provided a “workable prototype of the cellular changes that can cause cancer” (Luria, 1960, p. 679). Among the speakers at the symposium to explicitly discuss the analogies between bacteriophage lysogeny and tumor viruses was the Italian-American virologist Renato Dulbecco, an exponent of the American Phage group, only recently initiated into the world of


cancer viruses. Remarking specifically on the parallels between the types of changes that the phages and tumor viruses induced in their respective hosts, he cited the RSV-induced transformation of chicken fibroblast in culture as one “clear example of conversion of the cells by the virus genome, analogous to conversion caused by bacteriophage” (Dulbecco, 1960, p. 756). He reiterated the analogy shortly thereafter in a report of his own experiments on the in vitro effects of a mouse polyoma virus, proposing that “the absence or low level of virus production and the resistance to superinfection, are similar to the properties of lysogenic bacterial cultures and suggest that the integrated virus exists as provirus” (Vogt & Dulbecco, 1960, p. 368). Proof for the existence of the cancerous provirus was not long in appearing. Recognizing that that the active principle in transforming normal cells to tumor cells was likely the DNA of the virus (Baltimore, 2012), Dulbecco succeeded in quickly gathering evidence (switching experimental systems from the polyoma virus to a simian virus) that despite being able to transform cells, contained no autonomously replicating viruses, leading to the conclusion that the “viral DNA is integrated in the transformed cells” (Sambrook, Westphal, Srinivasan, & Dulbecco, 1968, p. 1291). In other words the virus had become a provirus. Dulbecco’s rapid turnaround success in this issue garnered him a share of the 1975 Nobel Prize in Physiology or Medicine for “discoveries concerning the interaction between tumour viruses and the genetic material of the cell” (Renato Dulbecco e Facts, 2013). Despite the fact that it had been touted as an example of a bonafide tumor-inducing virus at the 1960 symposium, the mechanism by which RSV could transform cells and induce tumors proved to be tougher nut to crack than the polyoma virus or SV40. This was because of a fundamental difference in its genetic makeup; RSV contained RNA not DNA, as it had been recently discovered (Crawford, 1960; Crawford & Crawford, 1961). This fact posed a problem in understanding the mechanisms of RSV-induced cell transformation, for the so-called “Central Dogma” of information flow in biologydfrom DNA to RNA to Proteindwas firmly in place by the time. The existence of RNA viruses was not in itself problematic because viruses were not cellular beings, but the idea that the RNA genomes could somehow become part of the host cell’s DNA was preposterous: “how could the Rous virus RNA integrate into cellular DNA? Therein lay the puzzle” (Baltimore, 1995, p. 1304). The answer to this question was provided by the work of HowardTemin a sometime student of Dulbecco who had focused his attention on RSV-induced transformation. Based on his initial experiments Temin, who would go on to share a portion of the same Nobel Prize as Dulbecco, suggested that that “the virus becomes equivalent to a cellular gene controlling cell morphology” (1960, p. 196). Being a purely genetic explanation that “contained no statement about the molecular nature of the provirus” (Fisher, 2009, p. 662), its implications were nothing short of heretical. As David Baltimoredthe third person to share in the1975 Prize with Temin and Dulbeccodwould explain later: Howard’s solution was chemically simple but without precedent: if the RNA were copied into DNA, then everything would fall into line. The RNA would become DNA, the DNA could integrate just like a lysogenic phage and the integrated genome could be transcribed back into RNA. Conceptually, a snap e but totally unacceptable to almost everyone then in molecular biology because it ran counter to the guiding dogma, that DNA makes RNA makes protein (Baltimore, 1995, p. 1304). Both the language and the immediate aftermath of Temin’s ideas, and for that matter Dulbecco’s, recall the hurdles faced by the Wollmans and Burnet to their initial proposals that suggested that

197

the bacteriophage was somehow turning into a gene during lysogeny. But by the 1960s there were more conceptual as well as technical tools at hand to test the question experimentally. Even though his claims were more controversial than those of Burnet and the Wollmans, Temin had the means to design experiments to test his theories. He conducted hybridization experiments which showed that the DNA of RSV-infected cells had regions that were homologous to the RNA genome of the virus (Temin,1964a) and thereby advanced his theory that RSV acted as “a carcinogenic agent by adding some new genetic information to the cell” (Temin, 1964b, p. 568). A few year later, both Temin and David Baltimoredworking independently of each other in different laboratoriesddiscovered reverse transcriptase, an enzyme capable of synthesizing DNA molecules using RNA as template (Baltimore, 1970; Temin & Mizutani, 1970), which decisively smashed the notion of a heresy to pieces. 5. Conclusion Temin’s resolution of the DNA provirus issue brings us full circle to where we began the storydi.e. the problem of the chicken sarcomadand signifies the end of the already desynchronized parallels in the story of the RSV and the bacteriophages. To claim any more similarities in the history of these two viruses would be no more legitimate than for any randomly picked pair of viruses. But the fact that these two viruses, or indeed any other entity we now consider a virus can be considered as such at all, is due to the merging of thought styles that enabled the very properties that prevented them from being considered as viruses to become those properties by which viruses came to be defined in the modern era. To consider the parallels in their earlier research trajectories therefore serves as a useful reminder to historians and scientists alike, to take into account, both the flux in scientific conceptsdas Löwy (1990, p. 88) reminds us, “we should not assume that [scientific] terms remain constant”das well as the broader scientific landscapes or styles that shape our practices and ideas. Acknowledgments The list of people to whom I owe immense gratitude for their help and encouragement in getting this paper out is very long indeed, but the first vote of thanks must go to our special issues editor and fellow contributor Robin Scheffler. It was through meeting Robin at the 2011 HSS meeting in Cleveland (thanks too to my PhD advisor Bill Summers for that introduction) that first spurred the idea for a session on viruses and cancer, which was realized in short order at the 2012 Three Society meeting in Philadelphia, where then Robin brought up the idea to turn the session into a special issue. Ever since, he worked tirelessly despite his own very busy schedule, to make sure that the rest of us have kept honest. The rest, both those who presented at that initial sessiondBrendan Clarke, Laura Stark & Doogab Yidand Alex Broadbent, who came on board this project later, are the next set of people to mention here. it’s been a great ride with all of you. To our commentators Angela Creager & Ton van Helvoort, and the editor-in-chief of this journal, Greg Radick, a special thank you for your support in putting this special issue together. On a more personal front, I would also like to extend my appreciation to the various peopledmost especially Lisa Onaga and Tim Fullerdwho patiently read and helped me to edit this paper through its multiple cycles. To Pierre-Olivier Methót and Bart Swiatczak, thank you both for your participation in a session which showed off the multiple contexts in which this work is relevant. The comments and ideas of the anonymous reviewers were enormously helpful and I certainly think the final paper is much improved as a result of my addressing

198


the issues they raised. As always all mistakes, omissions and warts are mine alone. References Ackerknecht, E. H. (1958). Historical notes on cancer. Medical History, 2(2), 114. Andrewes, C. H. (1950). The bearing of recent work on the virus theory of cancer. British Medical Journal, 1(4645), 81e85. Andrewes, C. H. (1960). Discussion of Dr. Luria’s paper. Cancer Research, 20(5 Part 1), 689e694. Andrewes, C. H. (1971). Francis Peyton Rous 1879e1970. Biographical memoirs of fellows of the Royal Society. (Great Britain), 17, 643e662. Avery, O. T., MacLeod, C. M., & McCarty, M. (1944). Studies on the chemical nature of the substance inducing transformation of pneumococcal types. The Journal of Experimental Medicine, 79(2), 137e158. Baltimore, D. (1970). Viral RNA-dependent DNA polymerase: RNA-dependent DNA polymerase in virions of RNA tumour viruses. Nature, 226(5252), 1209e1211. Baltimore, D. (1995). Thinking about Howard Temin. Genes and Development, 9, 1303e1307. Baltimore, D. (2012). Renato Dulbecco (1914e2012). Science, 335(6076), 1587e1587. http://dx.doi.org/10.1126/science.1221692. Bashford, E. (1905). Are the problems of cancer insoluble? British Medical Journal, 2(2345), 1507e1511. Becsei-Kilborn, E. (2003). Going against the grain: Francis Peyton Rous (1879e1970) and the search for the cancer virus (Ph.D.). Chicago: University of Illinois at Chicago Becsei-Kilborn, E. (2010). Scientific discovery and scientific reputation: The reception of Peyton Rous’ discovery of the chicken sarcoma virus. Journal of the History of Biology, 43(1), 111e157. Bordet, J. (1922). Concerning the theories of the so-called bacteriophage. British Medical Journal, 2(3216), 296. Bordet, J., & Ciuca, M. (1921). Remarques sur l’historique des recherches concernant la lyse microbienne transmissible. Comptes rendus des séances de la Société de Biologie et de ses filiales Paris, 84, 745e747. Burian, R. M., & Gayon, J. (1999). The French school of genetics: From physiological and population genetics to regulatory molecular genetics. Annual Review of Genetics, 33, 313e349. Burnet, F. M. (1926). Letter to Linda Druce, November 1926. F. M. Burnet Collection. University of Melbourne Archives. Series 2, File 10. Burnet, F. M. (1929). A method for the study of bacteriophage multiplication in broth. British Journal of Experimental Pathology, 10(2), 109e115. Burnet, F. M. (1930). Bacteriophage and cognate phenomena. A System of Bacteriology, 7, 463e509. Burnet, F. M. (1934). The bacteriophages. Biological Reviews and Biological Proceedings [of the Cambridge Philosophical Society], 9, 332e350. Burnet, F. M., & Lush, D. (1936). Induced lysogenicity and mutation of bacteriophage within lysogenic bacteria. The Australian Journal of Experimental Biology and Medical Science, 14, 27e38. Burnet, F. M., & McKie, M. (1929). Observations on a permanently lysogenic strain of B. entiritidis gaertner. The Australian Journal of Experimental Biology and Medical Science, 6(4), 277e284. Cohn, M. (1978). Burnet, lysogeny and creativity. Walter and Eliza Hall Institute of Medical Research Annual Review, 1978e9, 9e13. Crawford, L. V. (1960). A study of the Rous sarcoma virus by density gradient centrifugation. Virology, 12(2), 143e153. Crawford, L. V., & Crawford, E. M. (1961). The properties of Rous sarcoma virus purified by density gradient centrifugation. Virology, 13(2), 227e232. Creager, A. N. (2007). Adaptation or selection? Old issues and new stakes in the postwar debates over bacterial drug resistance. Studies in History and Philosophy of Biological and Biomedical Sciences, 38(1), 159e190. Creager, A. N. H. (2002). The life of a virus: Tobacco mosaic virus as an experimental model, 1930e1965. University of Chicago Press. Creager, A. N. H., & Gaudillière, J.-P. (2001). Experimental technologies and techniques of visualisation: Cancer as a viral epidemic, 1930e1960. In Heredity and infection: The history of disease transmission (pp. 203e260). New York and London: Routledge. D’Herelle, F. (1916). Sur un bacille dysentériques atypique. Annales de l’Institut Pasteur, 301, 145e147. D’Herelle, F. (1917a). Sur un microbe invisible antagoniste des bacilles dysentériques. Comptes rendus de l’Académie des Sciences, 165, 373e375. D’Herelle, F. (1917b). An invisible microbe that is antagonistic to the dysentery bacillus. In T. D. Brock (Trans.), 1999. Milestones in microbiology: 1546 to 1940 (pp. 157e159). Publisher unknown. D’Herelle, F. (1922a). The nature of bacteriophage. The British Medical Journal, 2(3216), 289e293. D’Herelle, F. (1922b). The bacteriophage: Its role in immunity (G. H. Smith, Trans.). Baltimore: Waverly Press D’Herelle, F. (1924). Immunity in natural infectious disease (authorized English translation by George H. Smith). Baltimore, MD: Williams and Wilkin. D’Herelle, F. (1931). An address on bacteriophagy and recovery from infectious diseases. Canadian Medical Association Journal, 24(5), 619e628. Duckworth, D. H. (1976). Who discovered bacteriophage? Microbiology and Molecular Biology Reviews, 40(4), 793.

Dulbecco, R. (1960). A consideration of virus-host relationship in virus-induced neoplasia at the cellular level. Cancer Research, 20(5, Part 1), 751e761. Dulbecco, R. (1976). Francis Peyton Rous. Biographical memoirs. National Academy of Sciences (U.S.), 48, 275e306. Ellis, E. L., & Delbrück, M. (1939). The growth of bacteriophage. The Journal of General Physiology, 22(3), 365e384. http://dx.doi.org/10.1085/jgp.22.3.365. Ewing, J. (October 6, 1935). Letter from James Ewing to Peyton Rous. Fisher, S. (2009). Not beyond reasonable doubt: Howard Temin’s provirus hypothesis revisited. Journal of the History of Biology, 43(4), 661e696. Fleck, L. (1979). Genesis and development of a scientific fact. In T. J. Trenn & F. Bradley, Trans., T. J. Trenn & R. K. Merton, (Eds.). University of Chicago Press (First published in German in 1935). Fujimura, J. H. (1987). Constructing “do-able” problems in cancer research: Articulating alignment. Social Studies of Science, 17(2), 257e293. Furth, J., & Metcalf, D. (1958). An appraisal of tumor-virus problems. Journal of Chronic Diseases, 8(1), 88e112. Galperin, C. (1987). Le bactériophage, la lysogénie et son déterminisme génétique. History and Philosophy of the Life Sciences, 9, 175e224. Gye, W. (1925). The aetiology of malignant new growths. Lancet, 206(5316), 109e117. Gye, W. E., & Andrewes, C. H. (1926). A study of the Rous fowl sarcoma No. 1: I. Filterability. British Journal of Experimental Pathology, 7(2), 81. Hughes, S. S. (1977). The virus: A history of the concept. Heinemann Educational Books. Kidd, J. G., Beard, J., & Rous, P. (1935). Certain factors determining the course of virus-induced tumors. Experimental Biology and Medicine, 33(1), 193e 195. Löwy, I. (1990). Variances in meaning in discovery accounts: The case of contemporary biology. Historical Studies in the Physical and Biological Sciences, 21(1), 87e121. Luria, S. E. (1960). Viruses, cancer cells, and the genetic concept of virus infection. Cancer Research, 20(5 Part 1), 677e688. Lwoff, A. (1953). Lysogeny. Bacteriological Reviews, 17(4), 269. Lwoff, A. (1957). The concept of virus. Journal of general microbiology, 17(2), 239e 253. Morange, M. (2005). What history tells us III. André Lwoff: From protozoology to molecular definition of viruses. Journal of Biosciences, 30(5), 591e594. Morris, H. (1903). The Bradshaw lecture on cancer and its origin. British Medical Journal, 2(2241), 1505e1511. Mößner, N. (2011). Thought styles and paradigms: A comparative study of Ludwik Fleck and Thomas S. Kuhn. Studies in History and Philosophy of Science Part A, 42, 416e425. Park, R. (1898). An inquiry into the etiology of cancer. With some reference to the latest investigations of the Italian pathologists. The American Journal of the Medical Sciences, 115(5), 503e519. Park, Roswell (1903). An epitome of the history of carcinoma. Medical Library and Historical Journal, 1(4), 239e250. Pasteur, L. (1890). La rage. Oeuvres de Pasteur, 6, 672e688. Peyrieras, N., & Morange, M. (2002). The study of lysogeny at the Pasteur Institute (1950e1960): An epistemologically open system. Studies in History and Philosophy of Biological and Biomedical Sciences, 33(3), 419e430. Plimmer, H. G. (1903). The parasitic theory of Cancer. British Medical Journal, 2(2241), 1511e1515. Renato Dulbecco e Facts. (2013). Nobelprize.org. Nobel Media AB. Accessed 17.02.14 . Rivers, T. M. (1932). The nature of viruses. Physiological Reviews, 12(3), 423. Rous, P. (1910). A transmissible avian neoplasm (sarcoma of the common fowl). Journal of Experimental Medicine, 12(5), 696e705. Rous, P. (1911). A sarcoma of the fowl transmissible by an agent separable from the tumour cells. Journal of Experimental Medicine, 13(4), 397e411. Rous, P. (1912). An avian tumour in its relation to the tumour problem. Proceedings of the American Philosophical Society, 51(205), 201e205. Rous, P. (1953). Letter to Christopher Andrewes, 21 April, 1953, from Peyton Rous Papers, The American Philosophical Society, quoted in E. Becsei-Kilborn, (2003). Going against the Grain: Francis Peyton Rous (1879e1970) and the search for the cancer virus (Ph.D.). Chicago: University of Illinois at Chicago Rous, P. (1958). Letter to Greer Williams, 4 November, 1958, from folder Williams, Greer, RG450. R762, Rous papers, Rockefeller University Archives, quoted in Van Helvoort, T. (2004). The start of a cancer research tradition: Peyton Rous, James Ewing, and viruses as the cause of cancer. In D. Stapleton (Ed.), Creating a tradition of biomedical research: Contributions to the history of the Rockefeller University (p. 206). New York: Rockefeller University Press. Rous, P. (1960). Opening remarks. Cancer Research, 20(5 Part 1), 672e676. Rous, P. (1967). The challenge to man of the neoplastic cell. Cancer Research, 27(11 Part 1), 1919e1924. Rous, P., & Lange, L. B. (1913). The characters of a third transplantable chicken tumour due to a filterable cause. A sarcoma of intracanalicular pattern. Journal of Experimental Medicine, 18(6), 651e664. Rous, P., & Murphy, J. B. (1914). On the causation by filterable agents of three distinct chicken tumours. The Journal of Experimental Medicine, 19(1), 52e69. Rous, P., Murphy, J. B., & Tytler, W. H. (1912). A filterable agent the cause of a second chicken-tumour, an osteochondrosarcoma. Journal of the American Medical Association, 59(20), 1793e1794.

N. Sankaran / Studies in History and Philosophy of Biological and Biomedical Sciences 48 (2014) 189e199 Sambrook, J., Westphal, H., Srinivasan, P. R., & Dulbecco, R. (1968). The integrated state of viral DNA in SV40-transformed cells. Proceedings of the National Academy of Sciences of the United States of America, 60(4), 1288e1295. Sankaran, N. (2006). Frank Macfarlane Burnet and the nature of the bacteriophage, 1924e1937 (Ph.D.). Yale University Sankaran, N. (2008). Stepping-stones to one-step growth: Frank Macfarlane Burnet’s role in elucidating the viral nature of the bacteriophages. Historical Records of Australian Science, 19(1), 83e100. Sankaran, N. (2010). Mutant bacteriophages, Frank Macfarlane Burnet, and the changing nature of “Genespeak” in the 1930s. Journal of the History of Biology, 43(3), 571e599. Stent, G. S. (2002). Prematurity in discovery. In E. B. Hook (Ed.), Prematurity in scientific discovery: On resistance and neglect (pp. 23e33). Berkeley & Los Angeles: University of California Press. Stent, G. S., Watson, J. D., & Cairns, J. (1966). Phage and the origins of molecular biology. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory of Quantitative Biology. Stowell, R. E. (1945). Thymonucleic acid in tumors. Cancer Research, 5(5), 283e294. Summers, W. C. (1999). Felix d’Herelle and the origins of molecular biology. New Haven: Yale University Press. Temin, H. M. (1960). The control of cellular morphology in embryonic cells infected with Rous sarcoma virus in vitro. Virology, 10(2), 182e197. Temin, H. M. (1964a). Homology between RNA from Rous sarcoma virus and DNA from Rous sarcoma virus-infected cells. Proceedings of the National Academy of Sciences of the United States of America, 52(2), 323. Temin, H. M. (1964b). Nature of the provirus of Rous sarcoma. Natl. Cancer Inst. Monogr, 17, 557e570. Temin, H. M., & Mizutani, S. (1970). RNA-dependent DNA polymerase in virions of Rous sarcoma virus. Nature, 226, 1211e1213. The International Cancer Conference. (1910). The British Medical Journal, 2(2599), 1266e1268. Twort, F. (1915). An investigation on the nature of ultra-microscopic viruses. Lancet, 186(4814), 1241e1243. Van Epps, H. L. (2005). Peyton Rous: Father of the tumour virus. The Journal of Experimental Medicine, 201(3), 320. Van Helvoort, T. (1991). What is a virus? The case of tobacco mosaic disease. Studies in History and Philosophy of Science, 22(4), 557e588. Van Helvoort, T. (1992a). Bacteriological and physiological research styles in the early controversy on the nature of the bacteriophage phenomenon. Medical History, 36(3), 243e270.

199

Van Helvoort, T. (1992b). The controversy between John H. Northrop and Max Delbrück on the formation of bacteriophage: Bacterial synthesis or autonomous multiplication? Annals of Science, 49(6), 545e575. Van Helvoort, T. (1993). Research styles in virus studies in the twentieth century: Controversies and the formation of consensus (Ph.D.). Maastricht: University of Limburg Van Helvoort, T. (1994a). The construction of bacteriophage as bacterial virus: Linking endogenous and exogenous thought styles. Journal of the History of Biology, 27(1), 91e139. Van Helvoort, T. (1994b). History of virus research in the twentieth century: The problem of conceptual continuity. History of Science, 32(96), 185e235. Van Helvoort, T. (1999). A century of research into the cause of cancer: Is the new oncogene paradigm revolutionary? History and Philosophy of the Life Sciences, 21(3), 293e330. Van Helvoort, T. (2004). The start of a cancer research tradition: Peyton Rous, James Ewing, and viruses as the cause of cancer. In D. Stapleton (Ed.), Creating a tradition of biomedical research: Contributions to the history of the Rockefeller University (pp. 191e209). New York: Rockefeller University Press. Varley, A. W. (1981). Early bacteriophage research: The contribution of Frank MacFarlane Burnet (M.Sc.). Lawrence, KS: University of Kansas Varley, A. W. (1986). Living molecules or autocatalytic enzymes: The controversy over the nature of bacteriophage 1915e1925 (Ph.D.). Lawrence, KS: University of Kansas, History Virchow, R. (1847). Zur Entwickelungsgeschichte des Krebses. Virchows Archiv, 1, 94e204. Vogt, M., & Dulbecco, R. (1960). Virus-cell interaction with a tumor-producing virus. Proceedings of the National Academy of Sciences of the United States of America, 46(3), 365. Waterson, A. P., & Wilkinson, L. (1978). An introduction to the history of virology. Cambridge University Press. Williams, W. R. (1908). The natural history of cancer, with special reference to its causation and prevention. William Wood and Co. Wollman, E. (1928). Bactériophagie et processus similaires. Hérédité ou infection? Bulletin de l Institut Pasteur, 26, 1e14. Wollman, E. (1935). The phenomenon of Twort-d’Herelle and it’s significance. Lancet, 226(5858), 1312e1314. Wollman, E., & Wollman, E. (1937). Les “phases” des bacteriophages (facteurs lysogènes). Comptes rendus des séances de la Société de Biologie et de ses filiales Paris, 124, 931e934.

Organization of shared and unshared sequences in the genomes of chicken endogenous and sarcoma viruses.

Archaeal viruses and bacteriophages: comparisons and contrasts.

Phosphoproteins of Rous sarcoma viruses.

Environmental bacteriophages: viruses of microbes in aquatic ecosystems.

The genome of the mammalian sarcoma viruses.

Chicken egg yolk enhances focus formation by subgroup B, C and D rous sarcoma viruses.

Oncogenicity of avian leukosis viruses of different subgroups and of mutants of sarcoma viruses.

In vitro isolation of stable rat sarcoma viruses.

Generation of new mouse sarcoma viruses in cell culture.

Murine xenotropic type C viruses. III. Phenotypic mixing with avian leukosis and sarcoma viruses.

Infection of chicken erythrocytes with influenza and other viruses.

Transformation of horse skin cells by type-C sarcoma viruses.

Structure of 50 to 70S RNA from Moloney sarcoma viruses.

Plaque assay of avian sarcoma viruses using casein.

Temperature-sensitive mutants of avian sarcoma viruses: genetic recombination between multiple or coordinate mutants and avian leukosis viruses.

Mallard or chicken? Comparing the isolation of avian influenza A viruses in embryonated Mallard and chicken eggs.

Viruses in the stools.

Vulpinic acids inhibit influenza (RNA) viruses but not herpes (DNA) viruses.

Intracellular restriction on the growth of induced subgroup E avian type C viruses in chicken cells.

env Gene of chicken RNA tumor viruses: extent of conservation in cellular and viral genomes.

Viruses and the Microbiota.

Viruses and the blood.

The origin of 7 S RNA in virions of avian sarcoma viruses.

Identification of morphological differences between avian influenza A viruses grown in chicken and duck cells.